IPsecme D. Migault
Internet-Draft Ericsson
Intended status: Standards Track M. Hatami
Expires: 18 August 2025 S. Céspedes
W. Atwood
Concordia University
D. Liu
Ericsson
T. Guggemos
LMU
C. Bormann
Universitaet Bremen TZI
D. Schinazi
Google LLC
14 February 2025
ESP Header Compression Profile
draft-ietf-ipsecme-diet-esp-05
Abstract
This document specifies Diet-ESP, a compression mechanisms for
control information in IPsec/ESP communications. The compression is
expressed through the Static Context Header Compression architecture.
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
This Internet-Draft will expire on 18 August 2025.
Copyright Notice
Copyright (c) 2025 IETF Trust and the persons identified as the
document authors. All rights reserved.
Migault, et al. Expires 18 August 2025 [Page 1]
Internet-Draft EHCP February 2025
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents (https://trustee.ietf.org/
license-info) in effect on the date of publication of this document.
Please review these documents carefully, as they describe your rights
and restrictions with respect to this document. Code Components
extracted from this document must include Revised BSD License text as
described in Section 4.e of the Trust Legal Provisions and are
provided without warranty as described in the Revised BSD License.
Table of Contents
1. Requirements notation . . . . . . . . . . . . . . . . . . . . 3
2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1. The Three compressors described in this specification . . 4
2.2. The scope of SCHC in this specification . . . . . . . . . 5
2.3. SCHC Rules and SCHC static context in this
specification . . . . . . . . . . . . . . . . . . . . . . 5
3. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 6
4. Diet-ESP Integration into the IPsec Stack . . . . . . . . . . 7
4.1. SCHC Parameters for Diet-ESP . . . . . . . . . . . . . . 10
4.2. Attributes for Rules Generation . . . . . . . . . . . . . 11
4.2.1. Compression/Decompression Actions in Diet-ESP . . . . 16
5. SCHC Compression for IPsec in Tunnel mode . . . . . . . . . . 17
5.1. Inner IP Compression (IIPC) . . . . . . . . . . . . . . . 18
5.1.1. Inner IP Payload Compression . . . . . . . . . . . . 18
5.1.2. Inner IPv6 Header Compression . . . . . . . . . . . . 18
5.1.3. Inner IPv4 Header Compression . . . . . . . . . . . . 20
5.2. ESP Payload Data Byte Alignment . . . . . . . . . . . . . 20
5.3. ESP Payload Data Byte Alignment . . . . . . . . . . . . . 21
5.4. Clear Text ESP Compression (CTEC) . . . . . . . . . . . . 21
5.5. Encrypted ESP Compression (EEC) . . . . . . . . . . . . . 22
6. SCHC Compression for IPsec in Transport mode . . . . . . . . 22
7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 22
8. Security Considerations . . . . . . . . . . . . . . . . . . . 23
9. Acknowledgements . . . . . . . . . . . . . . . . . . . . . . 25
10. References . . . . . . . . . . . . . . . . . . . . . . . . . 25
10.1. Normative References . . . . . . . . . . . . . . . . . . 25
10.2. Informative References . . . . . . . . . . . . . . . . . 27
Appendix A. Appendix . . . . . . . . . . . . . . . . . . . . . . 28
A.1. Illustrative Examples of Diet-ESP in Tunnel Mode . . . . 28
A.1.1. Json representation in Tunnel mode . . . . . . . . . 28
A.1.2. Attributes for Rule Generation (AfRG) . . . . . . . . 29
A.1.3. Diet-ESP Compression . . . . . . . . . . . . . . . . 29
A.1.4. Diet-ESP Decompression . . . . . . . . . . . . . . . 30
A.2. Illustrative Examples of Diet-ESP in Transport Mode . . . 30
A.2.1. Json representation in Transport mode . . . . . . . . 31
A.2.2. Attributes for Rule Generation (AfRG) . . . . . . . . 33
A.2.3. Inner IP Packet (IIP) . . . . . . . . . . . . . . . . 33
Migault, et al. Expires 18 August 2025 [Page 2]
Internet-Draft EHCP February 2025
A.2.4. Diet-ESP Compression . . . . . . . . . . . . . . . . 33
A.2.5. Compressed Packet Output . . . . . . . . . . . . . . 34
A.2.6. Diet-ESP Decompression . . . . . . . . . . . . . . . 34
A.2.7. GitHub Repository: Diet-ESP SCHC Implementation . . . 34
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 34
1. Requirements notation
The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and
"OPTIONAL" in this document are to be interpreted as described in
BCP 14 [RFC2119] [RFC8174] when, and only when, they appear in all
capitals, as shown here.
2. Introduction
The Encapsulating Security Payload (ESP) [RFC4303] protocol is part
of the IPsec [RFC4301] suite of protocols and can provide
confidentiality, data origin authentication, integrity, anti-replay,
and traffic flow confidentiality. The set of services ESP provides
depends on the Security Association (SA) parameters negotiated
between devices.
An ESP packet is composed of the ESP Header, the ESP Payload Data,
the ESP Trailer, and the Integrity Check Value (ICV). ESP has two
modes of operation: Transport and Tunnel. In Transport mode, the ESP
Payload Data consists of the payload of the original IP packet; the
ESP Header is inserted after the original IP packet header. In
Tunnel mode, commonly used for VPNs, the ESP Header is placed after
an outer IP header and before the inner IP packet headers of the
original datagram. This ensures both the original IP headers and
payload are protected. Consequently, the ESP Payload Data field
contains either the payload from the original IP packet or the fully-
encapsulated IP packet, in transport mode or tunnel mode,
respectively.
The ESP Trailer, placed at the end of the ESP Payload Data, includes
fields such as Padding and Pad Length to ensure proper alignment, and
Next Header to indicate the protocol following the ESP header. The
ICV, calculated over the ESP Header, ESP Payload Data, and ESP
Trailer, is appended after the ESP Trailer to ensure packet
integrity. For a simplified overview of ESP, readers are referred to
Minimal ESP [RFC9333].
While ESP is effective in securing traffic, compression can reduce
packet sizes, enhancing performance in networks with limited
bandwidth. In such environments, reducing the size of transmitted
packets is essential to improve efficiency. This document defines
Migault, et al. Expires 18 August 2025 [Page 3]
Internet-Draft EHCP February 2025
Diet-ESP, a protocol that includes different compression/
decompression (C/D) of various structures processed by ESP. These C/
D are expressed through the Static Context Header Compression and
Fragmentation (SCHC) framework [RFC8724]. The structure of the ESP
packet to be compressed is shown in Figure Figure 1.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ----
| Security Parameters Index (SPI) | ^Auth.
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |Cove-
| Sequence Number | |rage
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | ----
| Payload Data* (variable) | | ^
~ Higher Layer Message (transport) or IP datagram (tunnel) ~ | |
| | |Encr.
+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ |Cove
| | Padding (0-255 bytes) | |rage*
+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ | |
| | Pad Length | Next Header | v v
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ ------
| Integrity Check Value-ICV (variable) |
~ ~
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Figure 1: Top-Level Format of an ESP Packet
2.1. The Three compressors described in this specification
The document outlines the three compressors utilized in Diet-ESP,
which are detailed as follows:
1. Inner IP Compression (IIPC): This process pertains to the
compression and decompression of the IP packet protected by ESP.
For outbound packets, the ESP incorporates the compressed Inner
IP (IIP) into the ESP Data Payload (refer to Figure 1). In the
case of inbound packets, decompression occurs after the
compressed IIP is retrieved from the Data Payload within the
Clear Text ESP packet.
Migault, et al. Expires 18 August 2025 [Page 4]
Internet-Draft EHCP February 2025
2. Clear Text ESP Compression (CTEC): This process pertains to the
compression and decompression of the ESP segment that is destined
for encryption. This encompasses the Payload Data and the ESP
Trailer, which includes the Padding, Pad Length, and Next Header
fields, as illustrated in Figure 1. At this stage, only the
latter fields are eligible for compression. For outbound
packets, the ESP subsequently encrypts the compressed Clear Text
ESP. For inbound packets, decompression takes place following
the decryption process of the ESP.
3. Encrypted ESP Compression (EEC): This process pertains to the
compression and decompression of the Encrypted ESP packet (EE),
which consists of the ESP Header, the encrypted payload, and the
Integrity Check Value (ICV). Since neither the encrypted payload
nor the ICV can be compressed, only the ESP Header, specifically
the SPI and SN fields, are subject to compression.
2.2. The scope of SCHC in this specification
SCHC [RFC8724] offers a mechanism for header compression as well as
an optional fragmentation feature. This document utilizes SCHC as a
practical means to illustrate the capability to compress and
decompress a structured payload. It is important to note that any
elements of SCHC that pertain to aspects other than compression or
decompression, such as fragmentation, fall outside the purview of
this document. The reference to SCHC herein is solely for
descriptive purposes related to compression and decompression, and it
is not anticipated that the general SCHC framework will be integrated
into the ESP implementation. The structured payloads addressed in
this specification pertain to internal structures managed by ESP for
the processing of an IP packet. Consequently, the compression and
decompression processes outlined in this document represent
supplementary steps for the ESP stack in handling the ESP packet.
2.3. SCHC Rules and SCHC static context in this specification
SCHC facilitates the compression and decompression of headers by
utilizing a common context that may encompass multiple Rules. Each
Rule is designed to correspond with specific values or ranges of
values within the header fields. When a Rule is successfully
matched, the corresponding header fields are substituted with the
Rule ID and the Compression Residue, which contains the remaining
bits after compression.
In the context of IPsec, the process of encryption or decryption
between IPsec peers necessitates a common context known as a Security
Association (SA). More broadly, the SA encompasses all essential
parameters required by the Encapsulating Security Payload (ESP) to
Migault, et al. Expires 18 August 2025 [Page 5]
Internet-Draft EHCP February 2025
handle both inbound and outbound packets. It is important to note
that SAs are unidirectional. Furthermore, IPsec can link both
outbound and inbound IP packets to the SA through Traffic Selectors
(TS) or Security Parameters Index (SPI). This capability allows
IPsec to uniquely associate outbound and inbound packets with a
specific context (SA), which contains all pertinent information for
IPsec processing.
This document adopts a comparable methodology for compression and
decompression, ensuring that the SA includes all necessary parameters
to create the unique Rule applicable for compressing or decompressing
each structured payload. This guarantees that each SA is linked to a
single Rule, thereby allowing the Rule ID to be omitted. The Rule
associated with each structured payload is generated based on
specific parameters referred to in this document as Attributes for
Rule Generation (AfRG) (see Section 4.2 for a more detailed
description). These AfRGs are negotiated through IKEv2 [RFC7296],
and in such cases, they are likely already included in the SA. Any
additional missing AfRGs are negotiated via
[I-D.ietf-ipsecme-ikev2-diet-esp-extension].
3. Terminology
ESP Header Compression: A method to reduce the size of ESP headers
and trailer using predefined compression rules and contexts to
improve efficiency.
ESP Trailer: A set of fields added at the end of the ESP payload,
including Padding, Pad Length, and Next Header, used to ensure
alignment and indicate the next protocol.
Inner IP C/D (IIPC): Expressed via the SCHC framework, IIPC
compresses/decompresses the inner IP packet headers.
Clear Text ESP C/D (CTEC): Expressed via the SCHC framework, CTEC
compresses/decompresses all fields that will later be encrypted by
ESP, which include the ESP Data and ESP Trailer.
Encrypted ESP C/D (EEC): Expressed via the SCHC framework, EEC
compresses/decompresses ESP fields that will not be encrypted by
ESP.
Security Parameters Index (SPI): As defined in [RFC4301],
Section 4.1.
Sequence Number (SN): As defined in [RFC4303], Section 2.2.
Static Context Header Compression (SCHC): A framework for header
Migault, et al. Expires 18 August 2025 [Page 6]
Internet-Draft EHCP February 2025
compression designed for LPWANs, as defined in [RFC8724].
Static Context Header Compression Rules (SCHC Rules): As defined in
[I-D.ietf-schc-architecture]
RuleID: A unique identifier for each Rule part of the Set of Rules.
SCHC Parameters: A set of predefined values used for SCHC
compression and decompression, ensuring byte alignment and proper
packet formatting based on the SCHC profile.
SCHC MAX_PACKET_SIZE: The maximum size of a SCHC-compressed packet
that can be transmitted without fragmentation.
Traffic Selector (TS): A set of parameters (e.g., IP address range,
port range, and protocol) used to define which traffic should be
protected by a specific Security Association (SA).
It is assumed that the reader is familiar with other SCHC terminology
defined in [RFC8376], [RFC8724], and eventually
[I-D.ietf-schc-architecture].
4. Diet-ESP Integration into the IPsec Stack
Figure 2 depicts the incorporation of Diet-ESP within the IPsec
framework.
IPsec necessitates that both endpoints agree on a shared context
known as the Security Association (SA). This SA is established via
IKEv2 and encompasses all Attributes for Rules Generation (AfRG)
(refer to Section 4.2) essential for formulating the Rules for each
compressor, specifically the Inner IP packet Compressor (IIPC), the
Clear Text ESP Compressor (CTEC), and the Encrypted ESP Compressor
(EEC).
When an Inner IP packet (IIP) is received, IPsec identifies the SA
linked to that packet. The ESP then determines the IIPC Rule from
the AfRG contained within the SA and compresses the IIP packet (IIPC:
C {IIP}). Subsequently, ESP constructs the Clear Text ESP payload
(CTE). The CTEC Rule is derived from the AfRG of the SA, allowing
for the compression of the CTE (CTEC: C {C {IIP}, ET}, where ET
represents the ESP Trailer). The ESP encrypts the payload, computes
the Integrity Check Value (ICV), and forms the ESP Encrypted payload
(EE). The EE Rule is derived from the AfRG of the SA, and then
utilized to compress the EE. The resulting compressed ESP extension
is integrated into an IP packet and transmitted as outbound traffic.
Migault, et al. Expires 18 August 2025 [Page 7]
Internet-Draft EHCP February 2025
For inbound traffic, the endpoint extracts the Security Parameter
Index (SPI) from the compressed EE, along with any other selectors
from the packet, to conduct a lookup for the SA. As outlined in
Section 8, since the SPI is derived from a potentially compressed ESP
Header, there may be instances where the endpoint must explore
multiple options, potentially leading to several lookups or, in the
worst-case scenario, multiple signature verifications (see Section 8
for a more detailed discussion). Once the SA is retrieved, the ESP
accesses the AfRG to ascertain the EEC Rule and proceeds to
decompress the EE. The ESP verifies the signature prior to
decryption. Following this, the CTEC Rule is derived from the AfRG
of the SA, allowing for the subsequent decompression. Finally, ESP
extracts the Data Payload from the CTE packet, retrieves the IIPC
Rule from the AfRG of the SA, and decompresses the IIP.
Note that implementations MAY differ from the architectural
description but it is assumed that the outputs will be the same.
Migault, et al. Expires 18 August 2025 [Page 8]
Internet-Draft EHCP February 2025
Endpoint Endpoint
+------------------------+ +------------------------+
| Inner IP packet | | Inner IP packet |
+------------------------+ +------------------------+
========|=================================================^========
IPsec | |
+------------------------+ |
| SA lookup | |
+------------------------+ |
========|=================================================|========
ESP | |
| +-------------------------------------+ |
| | Security Association | |
| | - Attributes for Rule Generations | |
| +-------------------------------------+ |
| | Generation of the IIPC Rule, | |
| | CTEC Rule and EEC Rule | |
| +-------------------------------------+ |
| |
v |
+------------------------+ +------------------------+
| IIPC: C {IIP} | | IIPC: D {IIP} |
+------------------------+ +------------------------+
| Formation of | | Extraction of |
| Clear Text ESP | | Data Payload |
+------------------------+ +------------------------+
| CTEC: | | CTEC: |
| C {C {IIP}, ET} | | D {C {IIP}, ET} |
+------------------------+ +------------------------+
| Encryption | | Decryption |
+------------------------+ +------------------------+
| Formation of | | Parsing |
| Encrypted ESP | | Encrypted ESP |
+------------------------+ +------------------------+
| EEC: | | EEC: |
| C {EH, C {C {IIP}, ET} | | D {EH, C {C {IIP}, ET} |
+------------------------+ +------------------------+
| | SA lookup |
| +------------------------+
========|=================================================^========
| |
v |
Outbound Traffic Inbound Traffic
Migault, et al. Expires 18 August 2025 [Page 9]
Internet-Draft EHCP February 2025
Figure 2: SCHC Integration into the IPsec Stack. Packets are
described for IPsec in tunnel mode. C designates the Compressed
header for the fields inside. IIP refers to the Inner IP packet,
EH refers to the ESP Header, and ET refers to the ESP Trailer.
IIPC, CTEC and EEC respectively designate the Inner IP Compress,
the Clear Text ESP Compressor and the Encrypted ESP Compressor.
4.1. SCHC Parameters for Diet-ESP
The SCHC Payload [RFC8724] is always in the form:
0 1 2 3 4 5 6 7
+-+-+-+-+-+-+-+---------...----------+~~~~~~~~~+---------------+
| RuleID | Compression Residue | Payload | SCHC padding |
+-+-+-+-+-+-+-+---------...----------+~~~~~~~~~+---------------+
|-------- Compressed Header ---------| |-- as needed --|
Figure 3: SCHC Packet
The RuleID is a unique identifier for each SCHC rule. It is included
in packets to ensure the receiver applies the correct decompression
rule, maintaining consistency in packet processing. Note that the
Rule ID does not need to be explicitly agreed upon and can be defined
independently by each party. The RuleID in Diet-ESP is expressed as
1 byte and is always elided as Rules are uniquely determined for
compressors.
Other variables required for the C/D in Diet-ESP are the following:
Maximum Packet Size: MAX_PACKET_SIZE is determined by the specific
IPsec ESP configuration and the underlying transport, but it is
typically aligned with the network’s MTU. The size constraints
are optimized based on the available link capacity and negotiated
parameters between endpoints.
SCHC Padding: Padding in SCHC is used to align data to a specific
boundary (typically byte-aligned or 8-bit aligned) to meet the
requirements for encryption which only considers octets instead of
bits. The SCHC Padding is only used over the CTE as described in
Section 5.3.
SCHC padding is used solely to ensure byte alignment for the
Compressed Inner IP Packet (IIPC) before the ESP padding is applied.
This distinction is necessary because ESP Padding may be compressed
or omitted depending on the negotiated alignment. In this document,
we refer to this specific use of SCHC padding as Byte Alignment.
Migault, et al. Expires 18 August 2025 [Page 10]
Internet-Draft EHCP February 2025
4.2. Attributes for Rules Generation
The list of attributes for the Rules Generation (AfRG) is shown in
Table 1. These attributes are used to express the various
compressions that operate at the IIPC, CTEC, and EEC.
As outlined in Section 4, this specification does not detail the
process by which the AfRG are established between peers. Instead,
such negotiations are addressed in
[I-D.ietf-ipsecme-ikev2-diet-esp-extension]. However, the AfRG can
be classified into two distinct categories. The first category
encompasses AfRG that are negotiated through a specific IKEv2
extension tailored for the negotiation of AfRG linked to a particular
profile, the Diet-ESP profile in this context. The AfRG referenced
in Table 1 in this category are: the DSCP Compression/Decompression
Action (CDA) dscp_cda, the ECN CDA ecn_cda, the Flow Label CDA
flow_label_cda, the ESP alignment alignment, the ESP SPI Least
Significant Bits (LSB) esp_spi_lsb, and the ESP Sequence Number LSB
esp_sn_lsb.
The second category pertains to AfRG that are negotiated through
IKEv2 exchanges or extensions that are not specifically designed for
compression purposes. This category includes AfRG associated with
TS, as identified in Table 1, which are the TS IP Version
ts_ip_version, the TS IP Source Start ts_ip_src_start, the TS IP
Source End ts_ip_src_end, the TS IP Destination Start
ts_ip_dst_start, the TS IP Destination End ts_ip_dst_end, the TS
Protocol ts_proto, the TS Port Source Start ts_port_src_start, the TS
Port Source End ts_port_src_end, the TS Port Destination Start
ts_port_dst_start, and the TS Port Destination End ts_port_dst_end.
These AfRG are derived from the Traffic Selectors established through
TSi/TSr payloads during the IKEv2 CREATE_CHILD_SA exchange, as
described in [RFC7296], Section 3.13. The AfRG IPsec Mode designated
as ipsec_mode in Table 1 is determined by the presence or absence of
the USE_TRANSPORT_MODE Notify Payload during the CREATE_CHILD_SA
exchange, as detailed in [RFC7296], Section 1.3.1. The AfRG Tunnel
IP designated as tunnel_ip in Table 1 is obtained from the IP address
of the IKE messages exchanged during the CREATE_CHILD_SA process, as
noted in [RFC7296], Section 1.1.3. The AfRGs designated as ESP
Encryption Algorithm esp_encr and ESP Security Parameter Index (SPI)
esp_spi in Table 1 are established through the SAi2/SAr2 payloads
during the CREATE_CHILD_SA exchange, while the AfRG designated as ESP
Sequence Number esp_sn in Table 1 is initialized upon the creation of
the Child SA and incremented for each subsequent ESP message.
The ability to derive the IPPC Rules for the IIPC from the agreed
Traffic Selectors is indicated by the variable iipc_profile.
Migault, et al. Expires 18 August 2025 [Page 11]
Internet-Draft EHCP February 2025
+===================+=====================+===========+============+
| Variable | Possible Values | Reference | Compressor |
+===================+=====================+===========+============+
| iipc_profile | "iipc_diet-esp", | ThisRFC | N/A |
| | "iipc_uncompress" | | |
+-------------------+---------------------+-----------+------------+
| dscp_cda | "uncompress", | ThisRFC | IIPC |
| | "lower", "sa" | | |
+-------------------+---------------------+-----------+------------+
| ecn_cda | "uncompress", | ThisRFC | IIPC |
| | "lower" | | |
+-------------------+---------------------+-----------+------------+
| flow_label_cda | "uncompress", | ThisRFC | IIPC |
| | "lower", | | |
| | "generated", "zero" | | |
+-------------------+---------------------+-----------+------------+
| ts_ip_version | "IPv4-only", | RFC7296 | IIPC |
| | "IPv6-only" | | |
+-------------------+---------------------+-----------+------------+
| ts_ip_src_start | IPv4 or IPv6 | RFC7296 | IIPC |
| | address | | |
+-------------------+---------------------+-----------+------------+
| ts_ip_src_end | IPv4 or IPv6 | RFC7296 | IIPC |
| | address | | |
+-------------------+---------------------+-----------+------------+
| ts_ip_dst_start | IPv4 or IPv6 | RFC7296 | IIPC |
| | address | | |
+-------------------+---------------------+-----------+------------+
| ts_ip_dst_end | IPv4 or IPv6 | RFC7296 | IIPC |
| | address | | |
+-------------------+---------------------+-----------+------------+
| ts_proto | TCP, UDP, UDP-Lite, | RFC7296 | IIPC |
| | SCTP, ANY, ... | | |
+-------------------+---------------------+-----------+------------+
| ts_port_src_start | Port number | RFC7296 | IIPC |
+-------------------+---------------------+-----------+------------+
| ts_port_src_end | Port number | RFC7296 | IIPC |
+-------------------+---------------------+-----------+------------+
| ts_port_dst_start | Port number | RFC7296 | IIPC |
+-------------------+---------------------+-----------+------------+
| ts_port_dst_end | Port number | RFC7296 | IIPC |
+-------------------+---------------------+-----------+------------+
| dscp_list | list of DSCP | RFCYYYY | IIPC |
| | numbers | | |
+-------------------+---------------------+-----------+------------+
| alignment | "8 bit", "16 bit", | ThisRFC | CTEC |
| | "32 bit", "64 bit" | | |
+-------------------+---------------------+-----------+------------+
Migault, et al. Expires 18 August 2025 [Page 12]
Internet-Draft EHCP February 2025
| ipsec_mode | "Tunnel", | RFC4301 | CTEC |
| | "Transport" | | |
+-------------------+---------------------+-----------+------------+
| tunnel_ip | IPv4 or IPv6 | RFC4301 | CTEC |
| | address | | |
+-------------------+---------------------+-----------+------------+
| esp_encr | ESP Encryption | RFC4301 | CTEC |
| | Algorithm | | |
+-------------------+---------------------+-----------+------------+
| esp_spi | ESP SPI | RFC4301 | EEC |
+-------------------+---------------------+-----------+------------+
| esp_spi_lsb | 0-32 | ThisRFC | EEC |
+-------------------+---------------------+-----------+------------+
| esp_sn | ESP Sequence Number | RFC4301 | EEC |
+-------------------+---------------------+-----------+------------+
| esp_sn_lsb | 0-32 | ThisRFC | EEC |
+-------------------+---------------------+-----------+------------+
Table 1: Set of Variables to generate IIPC, CTEC and EEC Rules
in Diet-ESP
Any variable starting with "ts_" is associated with the Traffic
Selectors (TSi/TSr) of the SA. The notation is introduced by this
specification but the definitions of the variables are in [RFC4301]
and [RFC7296].
The Traffic Selectors may result in a quite complex expression, and
this specification restricts that complexity. This specification
restricts the resulting TSi/TSr to a single type of IP address (IPv4
or IPv6), a single protocol (e.g., UDP, TCP, or ANY), a single port
range for source and destination. This specification presumes that
the Traffic Selectors can be articulated as a result of
CREATE_CHILD_SA with only one Traffic Selector [RFC7296],
Section 3.13.1 in both TSi and TSr payloads (as described in
[RFC7296], Section 3.13). The TS Type MUST be either
TS_IPV4_ADDR_RANGE or TS_IPV6_ADDR_RANGE.
Let the resulting Traffic Selectors TSi/TSr be expressed via the
Traffic Selector structure defined in [RFC7296], Section 3.13.1. We
designate the local TS the TS - either TSi or TSr - sent by the local
peer. Conversely we designate as remote TS the TS - either TSi or
TSr - sent by the remote peer.
The details of each parameter are the following:
iipc_profile: designates the behavior of the IIPC layer. When set
to "iipc_uncompress" IIPC is not performed. This specification
describes IIPC that corresponds to the "iipc_diet-esp" profile.
Migault, et al. Expires 18 August 2025 [Page 13]
Internet-Draft EHCP February 2025
flow_label_cda: indicates the Flow Label CDA, that is how the Flow
Label field of the inner IPv6 packet or the Identification field
of the inner IPv4 packet is compressed / decompressed - See
Section 4.2.1 for more information. In a nutshell, "uncompress"
indicates that Flow Label (resp. Identification) are not
compressed. "lower" indicates the value is read from the outer IP
header - eventually with some adaptations when inner IP packet and
outer IP packets have different versions. "generated" indicates
that the field is generated by the receiving party. In that case,
the decompressed value may take a different value compared to its
original value. "zero" indicates the field is set to zero.
dscp_cda: indicates the DSCP CDA, that is how the DSCP values of the
inner IP packet are compressed / decompressed - See Section 4.2.1
for more information. In a nutshell, "uncompress" indicates that
DSCP are not compressed. "lower" indicates the value is read from
the outer IP header - eventually with some adaptations when inner
IP packet and outer IP packets have different versions. "sa"
indicates, compression is performed according to the DSCP values
agreed by the SA (dscp_list).
ecn_cda: indicates ECN CDA, that is how the ECN values of the inner
IP packet are compressed / decompressed - See Section 4.2.1 for
more information. In a nutshell, "uncompress" indicates that DSCP
are not compressed. "lower" indicates the value is read from the
outer IP header - eventually with some adaptations when inner IP
packet and outer IP packets have different versions.
ts_ip_version: designates the TS IP version. Its value is set to
"IPv4-only" when only IPv4 IP addresses are considered and to
"IPv6-only" when only IPv6 addresses are considered. Practically,
when IKEv2 is used, it means that the agreed TSi or TSr results
only in a mutually exclusive combination of TS_IPV4_ADDR_RANGE or
TS_IPV6_ADDR_RANGE payloads. If TS Type of the resulting TSi/TSr
is set to TS_IPV4_ADDR_RANGE, ts_ip_version takes the value
"IPv4-only". Respectively, if TS Type is set to
TS_IPV6_ADDR_RANGE, ts_ip_version is set to "IPv6-only".
ts_ip_src_start: designates the TS IP Source Start, that is the
starting value range of source IP addresses of the inner packet
and has the same meaning as the Starting Address field of the
local TS.
ts_ip_src_end: designates TS IP Source End that is the high end
value range of source IP addresses of the inner packet and has the
same meaning as the Ending Address field of the local TS.
ts_ip_dst_start: designates the TS IP Destination Start, that is the
Migault, et al. Expires 18 August 2025 [Page 14]
Internet-Draft EHCP February 2025
starting value range of destination IP addresses of the inner
packet and has the same meaning as the Starting Address field of
the remote TS.
ts_ip_dst_end: designates the TS IP Destination End, that is the
high end value range of destination IP addresses of the inner
packet and has the same meaning as the Ending Address field of the
remote TS.
ts_proto: designates the TS Protocol, that is the Protocol ID of the
resulting TSi/TSr. This profile considers the specific protocol
values "TCP", "UDP", "UDP-Lite", "SCTP" and "ANY". The
representation of "ANY" is given in [RFC4301], Section 4.4.4.2.
ts_port_src_start: designates the TS Port Source Start, that is the
the starting value of the source port range of the inner packet
and has the same meaning as the Start Port field of the local TS.
ts_port_src_end: designates the TS Port Source End, that is the high
end value range of the source port range of the inner packet and
has the same meaning as the End Port field of the local TS.
ts_port_dst_start: designates TS Port Destination Start, that is the
starting value of the destination port range of the inner packet
and has the same meaning as the Start Port field of the remote TS.
ts_port_dst_end: designates TS Port Destination End, that is the
high end value range of the destination port range of the inner
packet and has the same meaning as the End Port field of the
remote TS.
IP addresses and ports are defined as a range and compressed using
the Least Significant Bits (LSB). For a range defined by start and
end values, msb( start, end ) is defined as the function that returns
the Most Significant Bits (MSB) that remains unchanged while the
value evolves between start and end. Similarly, lsb( start, end ) is
defined as the function that returns the LSB that changes while the
value evolves between start and end. Finally, len( x ) is defined as
the function that returns the number of bits of the bit array x.
dscp_list: designates the list of DSCP values associated to the
inner traffic - see for example [I-D.mglt-ipsecme-dscp-np]. These
are not Traffic Selectors, but the compression mandates that the
packets take one of these listed DSCP values.
alignment: designates the ESP alignment as defined by [RFC4303].
ipsec_mode: designates the IPsec Mode defined in [RFC4301]. In this
Migault, et al. Expires 18 August 2025 [Page 15]
Internet-Draft EHCP February 2025
document, the possible values are "tunnel" for the Tunnel mode and
"transport" for the Transport mode.
tunnel_ip: designates the Tunnel IP address of the tunnel defined in
[RFC4301]. This field is only applicable when the Tunnel mode is
used. That IP address can be an IPv4 or IPv6 address.
esp_encr: designates the ESP Encryption Algorithm - also designated
as Transform 1 in [RFC7296], Section 3.3.2. The algorithm is
needed to determine whether the ESP Payload Data needs to be
aligned to some predefined block size and if the ESP Pad Length
and Padding fields can be compressed. For the purpose of
compression it is RECOMMENDED to use algorithms that already
compressed their IV [RFC8750].
esp_spi: designates the Security Parameter Index defined in
[RFC4301].
esp_spi_lsb: designates the LSB to be considered for the compressed
SPI. A value of 32 for esp_spi_lsb will leave the SPI unchanged.
esp_sn: designates the ESP Sequence Number (SN) field defined in
[RFC4301].
esp_sn_lsb: designates the LSB to be considered for the compressed
SN. It works similarly to ESP SPI LSB (see esp_spi_lsb).
4.2.1. Compression/Decompression Actions in Diet-ESP
In addition to the Compression/Decompression Actions (CDAs) defined
in [RFC8724], Section 7.4, this specification uses the CDAs presented
in Table 2. These CDAs are either a refinement of the compute- * CDA
or the result of a combined CDA.
Migault, et al. Expires 18 August 2025 [Page 16]
Internet-Draft EHCP February 2025
+========================+=============+======================+
| Action | Compression | Decompression |
+========================+=============+======================+
| lower | elided | Get from lower layer |
+------------------------+-------------+----------------------+
| generated (Flow Label) | elided | Compute flow label |
+------------------------+-------------+----------------------+
| checksum | elided | Compute checksum |
+------------------------+-------------+----------------------+
| ESP padding | elided | Compute padding |
+------------------------+-------------+----------------------+
| hop limit | elided | Get from lower layer |
+------------------------+-------------+----------------------+
| Byte Alignment | send | Compute padding |
+------------------------+-------------+----------------------+
Table 2: EHCP ESP related parameter
lower: is only used in a Tunnel Mode and indicates that the fields
of the inner IP packet header are generated from the corresponding
fields of the Tunnel IP header fields. This CDA can be used for
the DSCP, ECN, and IPv6 Flow Label (resp. IPv4 identification)
fields.
generated: indicates that a brand new Flow Label/Identification
field is generated following [RFC6437], [RFC6864].
checksum: indicates that a checksum is computed accordingly.
Typically, the checksum CDA has a different implementation for
IPv4, UDP, TCP,...
ESP padding: indicates that the ESP padding bytes are generated
accordingly.
hop limit: indicates that the hop limit is derived from the outer
IPv6 header.
Byte Alignment: indicates that the the SCHC Byte Alignment bits are
generated accordingly.
5. SCHC Compression for IPsec in Tunnel mode
Migault, et al. Expires 18 August 2025 [Page 17]
Internet-Draft EHCP February 2025
5.1. Inner IP Compression (IIPC)
When iipc_profile is set to "iipc_uncompress", the packet is
uncompressed. When iipc_profile is set to "iipc_diet-esp", IIPC
proceeds to the compression of the inner IP Packet composed of an IP
Header and an IP Payload. The compression of the inner IP Payload is
described in Section 5.1.1.
The IP Header is compressed when ipsec_mode is set to "Tunnel" and
left uncompressed otherwise. ts_ip_version determines how the IPv6
Header (resp. the IPv4 header) is compressed - see Section 5.1.2
(resp. Section 5.1.3).
5.1.1. Inner IP Payload Compression
The compression only affects UDP, UDP-Lite, TCP or SCTP packets and
the type of packet is determined by the IP header.
For UDP, UDP-Lite, TCP and SCTP packets, source ports, destination
ports, and checksums are compressed. For source port (resp.
destination port) only the least significant bits are sent. FL is
set to 16 bits, TV is set to msb( ts_port_src_start, ts_port_src_end
) ( resp. ts_port_dst_start, ts_port_dst_end ), MO is set to "MSB"
and CDA to "LSB". The checksum is elided, FL is set to 16 bits, TV
is not set, MO is set to "ignore" and CDA is set to "checksum". This
may result in decompressing a zero-checksum UDP packet with a valid
checksum, but this has no impact as a valid checksum is universally
accepted.
For UDP or UDP-Lite the length field is elided. FL is set to 16, TV
is not set, MO is set to "ignore".
5.1.2. Inner IPv6 Header Compression
The version field is elided, FL is set to 3, TV is set to
ts_ipversion, MO is set to "equal" and CDA is set to "not-sent".
Traffic Class is composed of the 6 bit DSCP and 2 bit ECN. The
compression of DSCP and ECN are defined independently.
DSCP values are compressed according to the dscp_cda value:
* If dscp_cda is set to "uncompress", the DSCP values are included
in the inner IP header. FL is set to 6 bits, TV is not set, MO is
set to "ignore", CDA is set to "sent-value".
* If dscp_cda is set to "lower", the DSCP field is elided and its
value is copied from the Tunnel IP header. FL is set to 6 bits,
TV is not set, MO is set to "ignore", CDA is set to "lower".
Migault, et al. Expires 18 August 2025 [Page 18]
Internet-Draft EHCP February 2025
* If dscp_cda is set to "sa", DSCP is compressed according to the
DSCP values of the SA. If dscp_list contains a single element,
the DSCP is elided, FL is set to 6 bits, TV is set to
dscp_list[0], MO is set to "equal" and CDA is set to "not-sent".
If dscp_list contains more than one DSCP value, FL is set to 6
bits, TV is set to dscp_list, MO is set to "match-mapping" and the
CDA is set to "mapping-sent". For ECN, FL is set to 2 bits, TV is
not set, MO is set to ignore and CDA is set to "value-sent".
ECN values are compressed according to the ecn_cda value:
* If ecn_cda is set to "uncompress", the ECN field is included in
the inner IP header. FL is set to 2 bits, TV is not set, MO is
set to "ignore", CDA is set to "sent-value".
* If ecn_cda is set to "lower", the ECN value is elided and the ECN
value is copied in the outer IP header. FL is set to 2 bits, TV
is not set, MO is set to "ignore", CDA is set to "lower".
Flow label is compressed according to the flow_label_cda value:
* If flow_label_cda is set to "uncompress", the Flow label is
included in the IPv6 Header. FL is set to 20 bits, TV is not set,
MO is set to "ignore", and CDA is set to "sent-value".
* If flow_label_cda is set to "lower", the Flow Label is elided and
read from the outer IP Header (See Section 4.2.1). FL is set to
20 bits, TV is not set, MO is set to "ignore", and CDA is set to
"lower". If the outer IP header is an IPv4 header, only the 16
LSB of the Flow Label are inserted into the IPv4 Header. At the
decompression, the 4 MSB of the Flow Label are set to 0.
* If flow_label_cda is set to "generated", the Flow Label is elided
and the Flow Label is then re-generated at the decompression (See
Section 4.2.1). The resulting Flow Label differs from the initial
value. FL is set to 20, TV is not set, MO is set to "ignore" and
CDA is set to "generated".
* If flow_label_cda is set to "zero", the Flow Label is elided and
set to 0 at decompression. A 0 value indicates no flow label is
present. Fl is set to 20 bits, TV is set to 0, MO is set to
"equal" and CDA is set to "not-sent".
Payload Length is elided and determined from the Tunnel IP Header
Payload Length as well as the decompressed Payload. FL is set to 16
bits, TV is not set, MO is set to "ignore", CDA is set to "lower".
Next Header is compressed according to ts_proto:
Migault, et al. Expires 18 August 2025 [Page 19]
Internet-Draft EHCP February 2025
* If ts_proto is the single value 0, Next Header is not compressed.
FL is set to 8 bits, TV is not set, MO is set to "ignore", CDA is
set to "sent-value".
* If ts_proto is a single non zero value, Next Header is compressed.
FL is set to 8 bits, TV is set to ts_proto, MO is set to "equal"
and CDA is set to "not-sent".
The IPv6 Hop Limit is read from the Tunnel IP Header Hop Limit. FL
is set to 8 bits, TV is not set, MO is set to "ignore" and CDA is set
to "lower."
The source and destination IPv6 addresses are compressed using MSB.
In both cases, FL is set to 128, TV is respectively set to
msb(ts_ip_src_start, ts_ip_src_ed) or msb(ts_ip_dst_start,
ts_ip_dst_end), the MO is set to "MSB," and the CDA is set to "LSB."
5.1.3. Inner IPv4 Header Compression
The fields Version, DSCP, ECN, Source Address and Destination Address
are compressed as described for IPv6 in Section 5.1.2. The field
Total Length (16 bits) is compressed similarly to the IPv6 field
Payload Length. The field Identification (16 bits) is compressed
similarly to the IPv6 field Flow Label. If the tunnel IP Header is
an IPv6 Header, the Identification is placed as the LSB of the IPv6
Header and the 4 remaining MSB are set to 0. The field Time to Live
is compressed similarly to the IPv6 Hop Limit field. The Protocol
field is compressed similarly to the last IPv6 Next Header field.
The Internet Header Length (IHL) is uncompressed, FL is set to 4
bits, TV is not set, MO is set to ignore and CDA is set to "value-
sent".
The IPv4 Header checksum is elided. FL is set to 16, TV is omitted,
MO is set to "ignore," and CDA is set to "checksum."
5.2. ESP Payload Data Byte Alignment
Deleted---see subsection below. SCHC operates on bits, and the
compression performed by SCHC may result in a bit payload that is not
aligned to a byte (8 bits) boundary. Protocols such as ESP expect
payloads to be aligned to byte boundaries (8-bit alignment). To
ensure this, we apply a padding by appending the SCHC_padding bits
and the SCHC_padding_len. SCHC_padding_len is encoded over 3 bits to
encode the number of bits that will compose the SCHC_padding field.
As a result SCHC_padding field has between 0 and 7 bits coded over
the SCHC_padding_len. The two fields are between 3 and 10 bits, so
if the complementing bits are less than or equal to 2 bits, the
Migault, et al. Expires 18 August 2025 [Page 20]
Internet-Draft EHCP February 2025
padding will result in adding an extra byte.
When the iipc_profile is set to "iipc_uncompress" there is no ESP
Payload Data Byte alignment. When iipc_profile is set to "iipc_diet-
esp" ESP Payload Data Byte Alignment is performed over the Compressed
Inner IP packet. This ensures that in both transport and tunnel
mode, the Payload Data later encrypted by ESP result in an integer
number of bytes.
5.3. ESP Payload Data Byte Alignment
SCHC operates on bits, and the compression performed by SCHC may
result in a bit payload that is not aligned to a byte boundary.
Protocols such as ESP expect payloads to be aligned to byte
boundaries (8-bit alignment). To ensure this, we apply a padding by
appending the Byte Alignment bits and the Byte Alignment Length
field. The Byte Alignment Length is encoded over 3 bits to indicate
the number of bits that will compose the Byte Alignment field. As a
result, the Byte Alignment field has between 0 and 7 bits, depending
on the required alignment. The total additional overhead can be up
to 10 bits (3-bit length field + 0-7 bits padding). If the
complementing bits are less than or equal to 2 bits, the padding will
result in adding an extra byte.
This Byte Alignment field is distinct from ESP Padding and is
required to ensure proper decryption without requiring additional
shifting operations in hardware. If the expected length of the
compressed residue is statically determinable based on the SA, the
padding length can be inferred, and the field may be omitted.
Otherwise, when the residue length depends on dynamic fields, the
length must be explicitly provided.
When the Byte Alignment is applied, it is performed only after the
IIPC stage and before the ESP Padding is added. The ESP Padding is
only compressed when alignment is explicitly set to 8 bits.
This ensures that in both transport and tunnel mode, the Payload Data
later encrypted by ESP results in an integer number of bytes.
5.4. Clear Text ESP Compression (CTEC)
Once the Inner IP Packet has undergone compression as outlined in
Section 5.1, followed by the ESP Byte Alignment detailed in
Section 5.3, the resulting compressed inner packet comprises a
specific number of bytes. To construct the Clear Text ESP Packet, it
is necessary to ascertain the ESP Payload Data, the Next Header, the
Pad Length, and the Padding fields.
Migault, et al. Expires 18 August 2025 [Page 21]
Internet-Draft EHCP February 2025
In transport mode, the IP header of the inner packet remains
uncompressed during the IIPC phase, and the ESP Payload Data is
derived from the inner packet. Conversely, in tunnel mode, the
Payload Data encompasses the entirety of the inner packet.
In transport mode, the Next Header field is obtained from either the
inner IP Header or the Security Association, as specified in
Section 5.1.3 or Section 5.1.2. In tunnel mode, the Next Header is
elided, as it is determined by ts_ip_version. FL is set to 8 bit, TV
is set to IPv4 or IPv6 depending on ts_ip_version, MO is set to
"equal" and CDA is set to "not-sent".
The ESP Pad Length and Padding fields are omitted only when ESP
alignment has been selected to "8bit" and esp_encr does not
necessitate a specific block size alignment, or if that block size is
one byte. This is represented by setting FL to (Pad Length + 1) * 8
bits, leaving TV unset, configuring MO to "ignore," and designating
CDA as padding. The ESP Padding and Pad Length may vary from their
decompressed counterparts. However, since the IPsec process removes
the padding, these variations do not affect packet processing. When
esp_encr requires a specific block size, the ESP Pad Length and
Padding fields remain uncompressed.
5.5. Encrypted ESP Compression (EEC)
SPI is compressed to its LSB. FL is set to 32 bits, TV is not set,
MO is set to "MSB( 4 - esp_spi_lsb)" and CDA is set to "LSB".
SN is compressed to their LSB, similarly to the SPI. FL is set to 32
bits, TV is not set, MO is set to "MSB( 4 - esp_sn_lsb)" and CDA is
set to "LSB".
6. SCHC Compression for IPsec in Transport mode
The transport mode mostly differs from the Tunnel mode in that the IP
header of the packet is not encrypted. As a result, the IP Payload
is compressed as described in Section 5.1.1. The IP header is not
compressed. The byte alignment of the Compressed Payload is
performed as described in Section 5.3. The Clear Text ESP
Compression is performed as described in Section 5.4 except for the
Next Header Field, which is compressed as described in Section 5.1.2.
7. IANA Considerations
We request the IANA to create a new registry for the IIPC Profile
Migault, et al. Expires 18 August 2025 [Page 22]
Internet-Draft EHCP February 2025
| IIPC Profile value | Reference |
+--------------------+-----------+
| "iipc_uncompress" | ThisRFC |
| "iipc_diet-esp" | ThisRFC |
We request IANA to create the following registries for the "diet-esp"
IIPC Profile.
| Flow Label CDA Value | Reference |
+----------------------+-----------+
| "uncompress" | ThisRFC |
| "generated" | ThisRFC |
| "lower" | ThisRFC |
| "zero" | ThisRFC |
| DSCP CDA Value | Reference |
+----------------------+-----------+
| "uncompress" | ThisRFC |
| "lower" | ThisRFC |
| "sa" | ThisRFC |
| ECN CDA Value | Reference |
+----------------------+-----------+
| "uncompress" | ThisRFC |
| "lower" | ThisRFC |
| Alignment | Reference |
+----------------------+-----------+
| "8 bit" | ThisRFC |
| "16 bit" | ThisRFC |
| "32 bit" | ThisRFC |
| "64 bit" | ThisRFC |
| IPsec mode Value | Reference |
+----------------------+-----------+
| "Tunnel" | ThisRFC |
| "Transport" | ThisRFC |
8. Security Considerations
The security considerations encompass those outlined in ESP [RFC4303]
as well as those pertaining to SCHC [RFC8724].
When employing ESP [RFC4303] in Tunnel Mode, the complete inner IP
packet is safeguarded against man-in-the-middle attacks through
cryptographic means, rendering it virtually impossible for an
attacker to alter any fields associated with either the inner IP
header or the inner IP payload. This level of protection is achieved
Migault, et al. Expires 18 August 2025 [Page 23]
Internet-Draft EHCP February 2025
by configuring the Flow Label CDA Value to "uncompress," the DSCP CDA
Value to either "uncompress" or "sa," and the ECN CDA Value to
"uncompress."
However, this protection is compromised if the Flow Label CDA Value,
DSCP CDA Value, or ECN CDA Values are set to "lower." In such cases,
the values from the inner packet for the respective fields will be
derived from the outer IP header, leaving these fields unprotected.
It is important to note that even the Authentication Header [RFC4302]
does not provide protection for these fields. When associated with a
CDA value of "lower," the level of protection is akin to that
provided in Transport mode. This vulnerability could be exploited by
an attacker within an untrusted domain, potentially disrupting load
balancing strategies that rely on the Flow Label [RFC6438][RFC6437].
More broadly, when the Flow Label CDA Value is set to "lower," the
relevant Flow Label Security Considerations from [RFC6437] apply.
Additionally, an attacker may manipulate the DSCP values to diminish
the priority of specific packets, resulting in packets from the same
flow having varying priorities, traversing different paths, and
introducing additional latency to applications, thereby facilitating
DDoS attacks. Consequently, these fields remain unprotected and are
susceptible to modification during transmission, which may also be
regarded as an acceptable risk.
When the Flow Label CDA Value is designated as "generated" or "zero,"
an attacker is unable to exploit the Flow Label field in any manner.
The inner packet received is anticipated to possess a Flow Label
distinct from that of the original encapsulated packet. However, the
Flow Label is assigned by the receiving gateway. When the value is
set to "zero," it is known, whereas when it is designated as
"generated," it must be produced in accordance with the
specifications outlined in [RFC6437].
The DSCP CDA Value is assigned as "sa" when DSCP values are linked to
Security Associations (SAs), but it should not be utilized when all
DSCP values are encompassed within a single SA. In such instances,
"uncompress" is recommended.
The encryption algorithm must adhere to the guidelines provided in
[RFC8221] to guarantee contemporary cryptographic protection.
The least significant bits (LSB) of the ESP Security Parameter Index
(SPI) determine the number of bits allocated to the SPI. An
acceptable quantity of LSB must ensure that the peer possesses a
sufficient number of SPIs, which is contingent upon the
implementation of the SA lookup employed. If a peer relies solely on
the SPI fields for SA lookup, then the number of LSB to consider must
be sufficiently large to include the number of SPIs. A peer may
Migault, et al. Expires 18 August 2025 [Page 24]
Internet-Draft EHCP February 2025
compress its SPIs differently, in which case a incoming packet may
have its SPI compressed to X bits while another packet may have its
SPI compressed to Y bits. The operator must be cognizant that if
multiple LSB values are permissible for each type of SA lookup, then
multiple SA lookups and signature verifications may be required. It
is advisable for a peer to ascertain the LSB associated with an
incoming packet in a deterministic manner.
The ESP SN LSB must be established in a manner that allows the
receiving peer to clearly ascertain the sequence number of the IPsec
packet. If this requirement is not met, it will lead to an invalid
signature verification, resulting in the rejection of the packet.
Furthermore, the LSB should have the capacity to accommodate the
maximum number of packets that may be in transit simultaneously.
This approach will guarantee that the last packet received is
correctly linked to the corresponding sequence number.
The ESP extension for IPv6 (and similarly for IPv4) requires a 64-bit
(or 32-bit) alignment. Choosing alternative alignment values may
result in a packet that fails to comply with this alignment
requirement, potentially leading to rejection. The necessity for
such alignment in IPv6 extensions arises from the fact that the
length field in an IPv6 header extension is defined in terms of
64-bit words, making proper alignment essential for accurate packet
parsing. Parsing of ESP does not present complications, as it is
compatible with IPv6; the ESP extension is processed exclusively by
the terminal IPsec peers and not by intermediary nodes. Furthermore,
the ESP extension lacks a dedicated length field. Instead, its
length is determined by the IPv6 Header Length field, which is
measured in bytes, along with the starting position of the ESP header
extension. Consequently, it remains entirely feasible to parse an
ESP extension that is not aligned to 64 bits. The same principle is
applicable to IPv4.
9. Acknowledgements
We would like to thank Laurent Toutain, Ana Caroline Minaburo and
Pascla Thubert for his guidance on SCHC. Robert Moskowitz for
inspiring the name "Diet-ESP" from Diet-HIP, Samita Chakrabart, Tero
Kivinen, Michael Richarson and Valery Smyslov for their long time
support. The authors would like to acknowledge the support from
Mitacs through the Mitacs Accelerate program.
10. References
10.1. Normative References
Migault, et al. Expires 18 August 2025 [Page 25]
Internet-Draft EHCP February 2025
[I-D.ietf-ipsecme-ikev2-diet-esp-extension]
Migault, D., Guggemos, T., Schinazi, D., Atwood, J. W.,
Liu, D., Preda, S., Hatami, M., and S. Cespedes, "Internet
Key Exchange version 2 (IKEv2) extension for Header
Compression Profile (HCP)", Work in Progress, Internet-
Draft, draft-ietf-ipsecme-ikev2-diet-esp-extension-03, 26
January 2025, <https://datatracker.ietf.org/doc/html/
draft-ietf-ipsecme-ikev2-diet-esp-extension-03>.
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>.
[RFC4301] Kent, S. and K. Seo, "Security Architecture for the
Internet Protocol", RFC 4301, DOI 10.17487/RFC4301,
December 2005, <https://www.rfc-editor.org/info/rfc4301>.
[RFC4303] Kent, S., "IP Encapsulating Security Payload (ESP)",
RFC 4303, DOI 10.17487/RFC4303, December 2005,
<https://www.rfc-editor.org/info/rfc4303>.
[RFC6437] Amante, S., Carpenter, B., Jiang, S., and J. Rajahalme,
"IPv6 Flow Label Specification", RFC 6437,
DOI 10.17487/RFC6437, November 2011,
<https://www.rfc-editor.org/info/rfc6437>.
[RFC6864] Touch, J., "Updated Specification of the IPv4 ID Field",
RFC 6864, DOI 10.17487/RFC6864, February 2013,
<https://www.rfc-editor.org/info/rfc6864>.
[RFC7296] Kaufman, C., Hoffman, P., Nir, Y., Eronen, P., and T.
Kivinen, "Internet Key Exchange Protocol Version 2
(IKEv2)", STD 79, RFC 7296, DOI 10.17487/RFC7296, October
2014, <https://www.rfc-editor.org/info/rfc7296>.
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/info/rfc8174>.
[RFC8221] Wouters, P., Migault, D., Mattsson, J., Nir, Y., and T.
Kivinen, "Cryptographic Algorithm Implementation
Requirements and Usage Guidance for Encapsulating Security
Payload (ESP) and Authentication Header (AH)", RFC 8221,
DOI 10.17487/RFC8221, October 2017,
<https://www.rfc-editor.org/info/rfc8221>.
Migault, et al. Expires 18 August 2025 [Page 26]
Internet-Draft EHCP February 2025
[RFC8724] Minaburo, A., Toutain, L., Gomez, C., Barthel, D., and JC.
Zuniga, "SCHC: Generic Framework for Static Context Header
Compression and Fragmentation", RFC 8724,
DOI 10.17487/RFC8724, April 2020,
<https://www.rfc-editor.org/info/rfc8724>.
[RFC8750] Migault, D., Guggemos, T., and Y. Nir, "Implicit
Initialization Vector (IV) for Counter-Based Ciphers in
Encapsulating Security Payload (ESP)", RFC 8750,
DOI 10.17487/RFC8750, March 2020,
<https://www.rfc-editor.org/info/rfc8750>.
10.2. Informative References
[I-D.ietf-schc-architecture]
Pelov, A., Thubert, P., and A. Minaburo, "Static Context
Header Compression (SCHC) Architecture", Work in Progress,
Internet-Draft, draft-ietf-schc-architecture-04, 6
February 2025, <https://datatracker.ietf.org/doc/html/
draft-ietf-schc-architecture-04>.
[I-D.mglt-ipsecme-dscp-np]
Migault, D., Halpern, J. M., Parkholm, U., and D. Liu,
"Differentiated Services Field Codepoints Internet Key
Exchange version 2 Notification", Work in Progress,
Internet-Draft, draft-mglt-ipsecme-dscp-np-01, 3 July
2024, <https://datatracker.ietf.org/doc/html/draft-mglt-
ipsecme-dscp-np-01>.
[OpenSCHC] "OpenSCHC a Python open-source implementation of SCHC
(Static Context Header Compression)", n.d.,
<https://github.com/openschc>.
[RFC4302] Kent, S., "IP Authentication Header", RFC 4302,
DOI 10.17487/RFC4302, December 2005,
<https://www.rfc-editor.org/info/rfc4302>.
[RFC6438] Carpenter, B. and S. Amante, "Using the IPv6 Flow Label
for Equal Cost Multipath Routing and Link Aggregation in
Tunnels", RFC 6438, DOI 10.17487/RFC6438, November 2011,
<https://www.rfc-editor.org/info/rfc6438>.
[RFC8376] Farrell, S., Ed., "Low-Power Wide Area Network (LPWAN)
Overview", RFC 8376, DOI 10.17487/RFC8376, May 2018,
<https://www.rfc-editor.org/info/rfc8376>.
Migault, et al. Expires 18 August 2025 [Page 27]
Internet-Draft EHCP February 2025
[RFC9333] Migault, D. and T. Guggemos, "Minimal IP Encapsulating
Security Payload (ESP)", RFC 9333, DOI 10.17487/RFC9333,
January 2023, <https://www.rfc-editor.org/info/rfc9333>.
Appendix A. Appendix
This appendix provides the details of the SCHC rules defined for
Diet-ESP compression, alongside an explanation and illustrative
examples for both Tunnel and Transport modes.
A.1. Illustrative Examples of Diet-ESP in Tunnel Mode
This section provides a structured example of how Diet-ESP operates
in Tunnel Mode. The example includes Attributes for Rule Generation
(AfRG), SCHC rules, the Inner IP packet (IIP), the compression
process, and the decompression process.
A.1.1. Json representation in Tunnel mode
In Tunnel Mode, the full inner IP packet is compressed before
encryption, making it more efficient for VPN scenarios. The
"iipc_diet-esp" profile is used to indicate compression of the inner
packet. The IPv6 Source and Destination Addresses are compressed
using "MSB", sending only the variable parts while keeping the most
significant bits. UDP Source and Destination Ports are compressed
with "LSB", reducing their size. "compute-length" removes the UDP
Length field, and "checksum" omits the UDP checksum, which is
recalculated at decompression. ESP SPI and Sequence Number are
compressed using "LSB" with an MSB mask. The "not-sent" CDA is used
for Next Header fields in both IPv6 and ESP, as their values are
predictable. ~~~json { "compressed": [ { "FID":
"IPv6.SourceAddress", "FL": 128, "TV": "2001:db8::1234", "MO": "MSB",
"CDA": "LSB" }, { "FID": "IPv6.DestinationAddress", "FL": 128, "TV":
"2001:db8::5678", "MO": "MSB", "CDA": "LSB" }, { "FID":
"IPv6.NextHeader", "FL": 8, "TV": 17, "MO": "equal", "CDA": "not-
sent" }, { "FID": "UDP.SourcePort", "FL": 16, "TV": 5001, "MO":
"MSB", "CDA": "LSB" }, { "FID": "UDP.DestinationPort", "FL": 16,
"TV": 4500, "MO": "MSB", "CDA": "LSB" }, { "FID": "UDP.Length", "FL":
16, "TV": null, "MO": "ignore", "CDA": "compute-length" }, { "FID":
"UDP.Checksum", "FL": 16, "TV": null, "MO": "ignore", "CDA":
"compute-checksum" }, { "FID": "ESP.SPI", "FL": 32, "TV": null, "MO":
"MSB(4 - esp_spi_lsb)", "CDA": "LSB" }, { "FID":
"ESP.SequenceNumber", "FL": 32, "TV": null, "MO": "MSB(4 -
esp_sn_lsb)", "CDA": "LSB" }, { "FID": "ESP.Padding", "FL": 8, "TV":
null, "MO": "ignore", "CDA": "padding" } ] } ~~~
Migault, et al. Expires 18 August 2025 [Page 28]
Internet-Draft EHCP February 2025
A.1.2. Attributes for Rule Generation (AfRG)
The SCHC rules for Tunnel Mode are derived from the following AfRG:
IPsec Mode: Tunnel
Traffic Selector IP Version: IPv6-only
Traffic Selector Source Address: 2001:db8::1234
Traffic Selector Destination Address: 2001:db8::5678
DSCP CDA: Lower
ECN CDA: Lower
ESP SPI Compression: LSB
ESP SN Compression: LSB
A.1.3. Diet-ESP Compression
The SCHC rules for the IIPC, CTEC, and EEC layers are defined as IIPC
to compresses IPv6 headers and UDP headers, CTEC to compresses ESP
Trailer fields and EEC to compresses ESP SPI and Sequence Number.
The IIPC original packet before compression consists of:
IPv6 Source Address: 2001:db8::1234
IPv6 Destination Address: 2001:db8::5678
UDP Source Port: 5001
UDP Destination Port: 4500
UDP Length: 16 bytes
ESP Payload Data
The compression process applies SCHC rules as follows:
A.1.3.1. IIPC: UDP Header Compression
UDP ports are compressed using the LSB technique.
UDP Length is removed (computed at decompression).
Migault, et al. Expires 18 August 2025 [Page 29]
Internet-Draft EHCP February 2025
UDP Checksum is omitted.
A.1.3.2. IIPC: IPv6 Header Compression
Source and Destination Addresses are compressed using MSB.
Next Header field is omitted.
A.1.3.3. CTEC: ESP Trailer Compression
Pad Length and Padding are omitted.
Next Header is omitted.
A.1.3.4. EEC: ESP Header Compression
SPI and SN are compressed using LSB.
Compressed Packet Output
The final compressed packet consists of the compressed ESP header,
IIPC compressed data, and payload.
A.1.4. Diet-ESP Decompression
The decompression process reverses these steps:
SPI and SN are reconstructed using the LSB values.
ESP Next Header and Padding fields are restored.
IPv6 Source and Destination Addresses are restored.
UDP ports are restored using the decompressed LSB values.
UDP Length and Checksum are recalculated.
A.2. Illustrative Examples of Diet-ESP in Transport Mode
This section follows the same structure as Tunnel Mode but applies to
Transport Mode, where the IP header remains uncompressed.
Migault, et al. Expires 18 August 2025 [Page 30]
Internet-Draft EHCP February 2025
A.2.1. Json representation in Transport mode
In Transport Mode, since the IP header remains uncompressed, only the
UDP payload and ESP fields are compressed. The new IANA-defined CDAs
are applied as follows: "checksum" is used for the UDP checksum,
meaning it is recalculated at decompression rather than transmitted.
"compute-length" eliminates the UDP Length field, as it can be
inferred from the payload size. "padding" ensures ESP padding aligns
with 8-bit boundaries. "not-sent" is applied to the IPv6 and ESP Next
Header fields because their values are predictable. The UDP Source
and Destination Ports use "LSB" compression, reducing overhead by
sending only the least significant bits. The ESP SPI and Sequence
Number are compressed similarly using "LSB" with an MSB mask. Since
the IP header is retained, the Source and Destination IPv6 Addresses
are fully transmitted using "sent-value".
[
{
"ipsec_mode": "Transport",
"ts_ip_version": "IPv6-only",
"compressed_fields": [
{
"FID": "IPv6.SourceAddress",
"FL": 128,
"TV": "2001:db8::1001",
"MO": "equal",
"CDA": "sent-value"
},
{
"FID": "IPv6.DestinationAddress",
"FL": 128,
"TV": "2001:db8::2002",
"MO": "equal",
"CDA": "sent-value"
},
{
"FID": "IPv6.NextHeader",
"FL": 8,
"TV": 17,
"MO": "equal",
"CDA": "not-sent"
},
{
"FID": "UDP.SourcePort",
"FL": 16,
"TV": 1234,
"MO": "MSB",
"CDA": "LSB"
Migault, et al. Expires 18 August 2025 [Page 31]
Internet-Draft EHCP February 2025
},
{
"FID": "UDP.DestinationPort",
"FL": 16,
"TV": 5678,
"MO": "MSB",
"CDA": "LSB"
},
{
"FID": "UDP.Length",
"FL": 16,
"TV": null,
"MO": "ignore",
"CDA": "compute-length"
},
{
"FID": "UDP.Checksum",
"FL": 16,
"TV": null,
"MO": "ignore",
"CDA": "checksum"
},
{
"FID": "ESP.SPI",
"FL": 32,
"TV": null,
"MO": "MSB(4 - esp_spi_lsb)",
"CDA": "LSB"
},
{
"FID": "ESP.SequenceNumber",
"FL": 32,
"TV": null,
"MO": "MSB(4 - esp_sn_lsb)",
"CDA": "LSB"
},
{
"FID": "ESP.Padding",
"FL": 8,
"TV": null,
"MO": "ignore",
"CDA": "padding"
},
{
"FID": "ESP.NextHeader",
"FL": 8,
"TV": 17,
"MO": "equal",
Migault, et al. Expires 18 August 2025 [Page 32]
Internet-Draft EHCP February 2025
"CDA": "not-sent"
}
]
}
]
A.2.2. Attributes for Rule Generation (AfRG)
The SCHC rules for Transport Mode are derived from the following
AfRG:
IPsec Mode: Transport
Traffic Selector IP Version: IPv6-only
Traffic Selector Source Address: 2001:db8::1001
Traffic Selector Destination Address: 2001:db8::2002
DSCP CDA: Lower
ECN CDA: Lower
ESP SPI Compression: LSB
ESP SN Compression: LSB
A.2.3. Inner IP Packet (IIP)
The original packet before compression consists of:
IPv6 Source Address: 2001:db8::1001
IPv6 Destination Address: 2001:db8::2002
UDP Source Port: 1234
UDP Destination Port: 5678
UDP Length: 18 bytes
ESP Payload Data
A.2.4. Diet-ESP Compression
Migault, et al. Expires 18 August 2025 [Page 33]
Internet-Draft EHCP February 2025
A.2.4.1. IIPC: UDP Header Compression
UDP ports are compressed using the LSB technique.
UDP Length is removed (computed at decompression).
UDP Checksum is omitted.
A.2.4.2. CTEC: ESP Trailer Compression
Pad Length and Padding are omitted.
Next Header is omitted.
A.2.4.3. EEC: ESP Header Compression
SPI and SN are compressed using LSB.
A.2.5. Compressed Packet Output
The final compressed packet consists of the compressed ESP header,
IIPC compressed data, and payload.
A.2.6. Diet-ESP Decompression
The decompression process mirrors the compression steps, restoring
SPI, SN, UDP headers, ESP Next Header, and Padding fields.
A.2.7. GitHub Repository: Diet-ESP SCHC Implementation
The source code for the implementation of the Diet-ESP profile,
including the compression and decompression logic using the SCHC
rules, is available on GitHub. Access the code at the following
link:
GitHub Repository: Diet-ESP SCHC Implementation
(https://github.com/mglt/pyesp/tree/master/examples/draft-diet-
esp.py)
This repository contains the rule definitions, examples, and source
code for implementing and testing the Diet-ESP profile. Refer to the
README file for setup instructions and usage guidelines.
Authors' Addresses
Daniel Migault
Ericsson
Email: daniel.migault@ericsson.com
Migault, et al. Expires 18 August 2025 [Page 34]
Internet-Draft EHCP February 2025
Maryam Hatami
Concordia University
Email: maryam.hatami@mail.concordia.ca
Sandra Céspedes
Concordia University
Email: sandra.cespedes@concordia.ca
J. William Atwood
Concordia University
Email: william.atwood@concordia.ca
Daiying Liu
Ericsson
Email: harold.liu@ericsson.com
Tobias Guggemos
LMU
Email: guggemos@nm.ifi.lmu.de
Carsten Bormann
Universitaet Bremen TZI
Email: cabo@tzi.org
David Schinazi
Google LLC
Email: dschinazi.ietf@gmail.com
Migault, et al. Expires 18 August 2025 [Page 35]